You can certainly use an external power supply.
If this target device is a product under development, it’s likely worth the effort sometime soon for you or someone else on the team to figure out what is going on. Nothing about the JS220 should cause this behavior, which likely indicates that the target device is marginal. It’s better to figure out why sooner before having lots of deployed units that start exhibiting this same issue!
Here are some questions to consider:
- Is that ~5 millisecond cycle time in the zoomed in screenshot your device rebooting?
- Is the brief drop of the LDO input to 3.1 V enough to brown-out something on the device?
- If you short JS220 I+ to I- without changing anything else, do you still see this problem? This bypasses the JS220’s internal impedance but keeps the wiring the same.
- Does the design follow all layout guidelines for any & all DC-DC regulators? For example, see the “PRINTED CIRCUIT BOARD (PCB) LAYOUT CONSIDERATIONS” in the ADP7156 datasheet. The fact that the input capacitor is on another sheet makes me concerned that it is not routed close to the ADP7156. Also, the addition of the 0 Ohm R8 can make layout worse.
- Did the design correctly derate these capacitors for working voltage? Are they still within spec?
- If you use an oscilloscope to measure the VIN and VOUT voltages during startup, how does it compare to battery only vs battery + JS220?
I hope the JS220 helps you create something great!